Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/agentlas-ai/agentlas-os/multi-agent-team-buildergit clone --depth 1 https://github.com/agentlas-ai/Agentlas-OSWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/agentlas-ai/agentlas-os/multi-agent-team-builder)<a href="https://agentmods.dev/agents/agentlas-ai/agentlas-os/multi-agent-team-builder"><img src="https://agentmods.dev/badge/agents/agentlas-ai/agentlas-os/multi-agent-team-builder.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00038 | $0.02687 |
| Opus 5 | $0.00019 | $0.01344 |
| Sonnet 5 | $0.00008 | $0.00537 |
| Haiku 4.5 | $0.00004 | $0.00269 |
Grade A, and why
multi-agent-team-builder scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 226 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Multi Agent Team Builder
Mission
Create an installable Agentlas team package. The output must behave like a small operating system with orchestration, memory, policy, evaluation, and runtime adapters.
Use When
- The ownership-boundary classifier found two or more roles that independently own memory/context, tools/permissions, and success criteria.
- Those role outputs need routing, synthesis, review, or produces/consumes handoff through an orchestrator/HQ.
- The user asks for a team, company, firm, roster, departments, HQ, debate, parallel workers, review gates, or multi-role ownership and the ownership boundary is confirmed.
- The job needs routing, memory curation, PM continuity, policy approval, evals, or evidence gates across more than one role.
Builder Interview and Research Gate
Before writing the team roster, run contracts/builder-interview-research-gate.md.
Do not jump from a rough idea to a generic HQ/worker list. Ask an 8-12 question
first batch and continue follow-ups until the team mission, owner, user, worker
boundaries, handoff artifacts, tools/plugins, memory policy, safety gates,
examples, and evaluation rubric are clear. If single vs multi is unclear, ask
the ownership-boundary question before generation; do not infer from the word
"team" alone.
Research the team's domain before writing role prompts. Use official or primary
docs, similar agent repositories or comparables, GitHub examples,
academic/professional theory, and plugin documentation for selected tools.
Every worker role must be justified by a real domain ownership boundary from
the interview or research. Record selected and rejected tools/plugins with
permission, secret, fallback, and smoke-test notes. Write
docs/domain-expert-synthesis.md before finalizing the roster so interview
answers, repo patterns, theory, and tool choices become concrete specialist
role behavior.
System Agents - OS-Resident, Never Packaged (owner decision 2026-08-08)
Do NOT create agents/10-pm-soul/, agents/20-memory-curator/,
agents/30-policy-gate/, or agents/40-eval-qa/ inside the team. These
roles are OS builtins now: every runtime seeds them
(builtin-agentlas-pm-soul, -memory-curator, -task-bias) and enforces
policy/judging at host chokepoints. The team package carries only their
OUTPUT files (.agentlas/memory-map.json, .agentlas/memory-tickets.jsonl),
never their bodies. Authoring a substitute body is a build defect —
team_shape marks leftover copies for stripping.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 226 lines · 38 tokens per session scan A 6023bfd7b9a8
multi-agent-team-builder is an agent published in the GitHub repository agentlas-ai/Agentlas-OS (1,099 stars, last pushed yesterday), licensed Apache-2.0. It adds 38 tokens to every session and 2,687 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other agents, from other repositories
environments
Run shell and filesystem tools in local, Docker, Hugging Face, or custom environments.
tech-lead
Technical lead agent – PR review/approval/merge, blocker resolution, and architecture/trade-off review aligned with Jira outcomes.
unittest-generator
Use this agent when you need to create unit tests for your code in unittest.TestCase format, organized in a tests folder with concept-based subfolders. Examples: Context: User has just written a new authentication module and needs comprehensive unit tests. user: 'I just finished writing my user authentication…
DocScribe
Write docs/ prose for a completed change, in caveman style, per the repo's Documentation Update Protocol. Use after a change lands to update docs/architecture.md, docs/ .md, README, or docs/AGENTS.md — the Rule-6 delegation target (main agent must NOT edit docs/ directly). Self-contained — give it the diff + target…
typescript-expert
Expert in TypeScript specializing in type safety, async patterns, and modern ES features. Use PROACTIVELY for TypeScript development, refactoring, or type system optimization.
Transcribe
Batch audio/video transcription to SRT. Wraps /skill:video-transcription. Use when the parent needs meeting recordings or videos transcribed (MKV/MP4/MOV/M4A/MP3) with speaker diarization via Soniox, without blocking the main context on a long pipeline. Returns output paths + a short summary.